01
Project
Create or select the extraction project.
Keep source documents and saved extraction runs within one project scope.
- Scope
- Named project
- Files
- Saved source documents
- History
- Extraction runs
MAG OptiAIOperational AIExtractAIEarly Access
Extract Document Information. Review the Structure.
Turn a business document into structured output you can inspect. Review extracted fields, quality notes and reading limits, then download JSON or available raw text with the source and run context in view.
Prepare the source. Run extraction. Check the output before a handoff.
Prefer to explore first? Start free14-day trial · lower limits. Beta access requires approval.
Built for document operations teams
Source documents and structured output
Cloud-based Pro Console
Guided evaluation
The document-extraction challenge
A Structured Result Needs Review and Context.
Source quality, reading coverage and output review shape a useful extraction handoff.
Challenge 01 / 04
Manual Re-Entry Takes Time and Attention
Useful information is often still inside a document.
Forms, delivery records and other business files can require repeated reading and copying. ExtractAI produces structured output for review so your team can inspect the returned information before a handoff.
Impact
Move from document reading to a structured result you can check.
The challenge is more than producing fields.
It is keeping the source, returned structure and review context connected.
Why MAG OptiAI
A Clearer Way to Extract and Review Document Information.
Keep the source document, extraction run and returned structure connected throughout the review.
A Defined Source and Run
Keep the document, extraction job and saved result within a named project.
Structured Output to Inspect
Review a field preview and the complete returned JSON before using the information.
A Clear Review Handoff
Check quality and reading limits, then download the available outputs.
Your source and extraction result stay connected.
Saved extraction runs retain the document and result context. Open a run intentionally to review its fields, quality notes and available outputs.
Process
How We Work
From a source document to structured output you can review.
MAG OptiAI guides your team through document preparation, extraction and result review.
02
Document
Prepare the document you want to extract.
Upload a document draft or review a saved document attached to the project.
- Input
- A supported document file
- Context
- Source name and format
- Review
- Document readiness
03
Extract
Run extraction and monitor its progress.
Queue the document extraction job and follow its lifecycle to a result or reported issue.
- Run
- Document extraction job
- Monitor
- Processing state
- Outcome
- Returned output or issue
04
Results
Review the structure and its source context.
Field preview · structured JSON · quality notes · reading limits · available downloads
- Inspect
- Fields and full JSON
- Check
- Quality and input coverage
- Keep
- Saved run and downloads
Core capabilities
What ExtractAI Can Do for Your Organization
A connected workflow for source documents, extraction jobs and structured result review.
Organize Extraction Projects
Keep source documents and saved extraction history within a named project.
Work With Supported Formats
Prepare PDF, office, text and image files through the supported upload workflow.
Run Document Extraction
Queue an extraction job and monitor its processing state.
Your document.
Structure to review.
Inspect Quality Notes
Review reported issues, parsing problems and fallback shaping where present.
Understand Reading Coverage
Check representation and input statistics for truncation or page limits.
Return to Saved Results
Open a saved extraction run intentionally and inspect its source context.
Review Extracted Fields
Inspect the top-level field preview and full structured JSON.
Download Available Outputs
Download structured JSON and raw-text previews where available.
Organize Extraction Projects
Keep source documents and saved extraction history within a named project.
Work With Supported Formats
Prepare PDF, office, text and image files through the supported upload workflow.
Run Document Extraction
Queue an extraction job and monitor its processing state.
Review Extracted Fields
Inspect the top-level field preview and full structured JSON.
Inspect Quality Notes
Review reported issues, parsing problems and fallback shaping where present.
Understand Reading Coverage
Check representation and input statistics for truncation or page limits.
Return to Saved Results
Open a saved extraction run intentionally and inspect its source context.
Download Available Outputs
Download structured JSON and raw-text previews where available.
All in one place.
The tools to prepare, extract and review in one connected workflow.
ExtractAI in action
See the Structure Alongside Its Source.
Review a source document alongside a returned field preview. This illustration explains the concept; it is not an actual extraction result.
Extracted fields deserve source review before use.
Inspect the field preview and full JSON alongside the document representation, quality notes and available raw-text preview.
Check source coverage and reported issues before downloading the output for your next review step.
Source document and structured outputIllustrative
Source
Delivery note
Reference: DN-1042
Items delivered: 12
delivery-note.pdf
Field preview
- Document type
- Delivery note
- Reference
- DN-1042
- Item count
- 12
Structured JSON · example
{
"document_type": "delivery_note",
"reference": "DN-1042",
"item_count": 12
}Who it’s for
Is ExtractAI a Good Fit for Your Team?
For teams turning business documents into information they can inspect and hand over.
Document Operations Teams
Prepare supported source documents and inspect the structured output from extraction.
Reviewers Checking Output Quality
Check field values, full JSON, quality notes and reading limits against the source.
Teams Preparing a Data Handoff
Open saved runs and download available outputs while keeping the source context in view.
Every document and extraction task is different. Evaluate the returned structure and coverage against your intended use.
The ExtractAI Evaluation
Real Problem. Real Evaluation. Real Impact.
Bring your extraction task and evaluate the fit with your organization’s source files.
What You Get
Access for a Larger Evaluation
Apply through the existing beta process to evaluate beyond the smaller free-trial limits.
A Guided Document Setup
Frame the extraction task and prepare a representative source document.
A Structured Review Workflow
Run extraction and inspect fields, JSON, quality notes and source coverage.
A Direct Feedback Channel
Discuss document preparation, unclear outputs and review requirements.
Outputs You Can Inspect
Evaluate the returned structure and available downloads against your intended handoff.
What We Ask From You
A Defined Extraction Task
Describe the document information your team needs to review or hand over.
Representative Source Files
Bring documents you are authorized to use. Redacted or anonymized files should preserve the structure needed for extraction.
Practical Feedback
Tell us where preparation, output quality or review needs improvement.
A Review of the Output
Check extracted information against the source and reported coverage before operational use.
Real Problem
Real Evaluation
Better informed review.
A real-world evaluation with your organization.
Bring your document-extraction task and representative files you are authorized to use. We will review the fit and help you get started.
ExtractAI Beta
Frequently Asked Questions
What to know before evaluating your document-extraction requirements.
What does ExtractAI produce?
The workflow returns structured document output, a field preview and result context. Inspect the complete JSON and quality notes rather than assume the returned structure is already verified.
Which document formats can I upload?
The current picker accepts PDF, text, Markdown, DOCX, RTF, HTML, PPTX, XLSX and supported images such as PNG, JPEG, WebP and TIFF. Format acceptance does not guarantee every file can be processed; check current workspace requirements and any reported issues.
Can I define a custom extraction schema here?
The current workflow takes a source document and returns structured output. It does not expose a custom schema or template editor. Review the returned fields and JSON against your task.
Does extraction always read the entire file?
No. Input statistics can report text truncation or page limits. Inspect the source representation and coverage notes before relying on the output.
What does a quality or fallback note mean?
It can indicate that the output needs checking, such as parsing issues, fallback shaping or partial source coverage. A completed run is not a guarantee that every extracted value is accurate or complete.
Can I reopen a saved extraction?
Yes. Open a saved extraction run from project history intentionally, then inspect its result and provenance. The source document and run remain part of the project context.
How does Early Access relate to the trial and beta?
Early Access describes product maturity. Start Free uses the existing 14-day trial with lower limits. Larger evaluations use the existing beta application and team approval; there is no separate access-request system.
What can I download?
Download the structured JSON when it is available. A raw-text preview can also be downloaded when the result contains one; it should not be assumed to reproduce the entire source document.
How do I apply for a larger evaluation?
Use Apply for the ExtractAI Beta to reach the existing product-specific application. Our team reviews your document requirements and confirms the fit for an evaluation.
Still have a question?
Talk to us about your document evaluation.
Ready to get started?
Ready to Evaluate ExtractAI with Your Organization’s Documents?
Tell us about your source documents and extraction requirements. We will review your application and confirm whether the current ExtractAI Beta is a good fit for your organization.
No payment requiredApplication reviewed by our team